The common world, in Citton's framework (drawing on Hannah Arendt), is the intersubjective reality that multiple minds inhabit together—not the physical world of objects but the meaningful world of shared references, common concerns, and mutually intelligible symbols. It is the world in which strangers can converse because they attend to the same news, argue because they recognize the same problems, and coordinate because they operate within the same framework of cultural meaning. The common world is not given—it is constructed, continuously and effortfully, through the practices and technologies that enable joint attention. When joint attention dissolves, the common world thins—becomes less substantial, less capable of supporting the weight of democratic institutions, cultural production, and mutual understanding. The thinning is not a metaphor: it describes a measurable reduction in shared reference, a statistical decline in the probability that any two randomly selected members of a society have attended to the same content, and a corresponding increase in the incomprehension and distrust that arise when people realize they no longer inhabit the same informational reality.
Citton's diagnosis of the common world's evaporation identifies AI as the accelerant, not the origin, of a process underway for decades. Cable television fragmented the broadcast audience—instead of three networks, three hundred channels, each serving a narrower demographic. The internet fragmented further—instead of channels, websites; instead of schedules, on-demand; instead of mass audiences, niche communities. Social media introduced algorithmic curation—instead of everyone seeing the same feed, each user seeing a personalized stream. Each step reduced the overlap between any two individuals' informational diets. Each step made the common world thinner. AI completes the dissolution by removing the final constraint: supply. Even personalized feeds had to draw from a bounded pool of human-created content, which meant some objects (viral videos, major news events, cultural phenomena) achieved sufficient penetration to function as shared referents. AI removes the supply constraint by generating content on-demand. The pool is no longer bounded. The shared object is no longer statistically likely.
The political consequences are already visible in phenomena that mystify observers lacking Citton's framework. The inability of democratic publics to agree on basic facts is not primarily about epistemic disagreement—it is about attentional fragmentation. Citizens operating in different algorithmic bubbles do not merely interpret the same evidence differently; they attend to different evidence, curated by systems optimizing for engagement rather than for truth or for the construction of common ground. The result is not pluralism (multiple perspectives on a shared reality) but what Citton calls attentional secession—the departure of subpopulations into separate informational realities that share progressively fewer common referents. The society fragments not through any coordinated separatist movement but through the aggregate effect of individually rational media consumption in an environment that has eliminated the structures forcing shared attention.
Citton's concept of synthetic joint attention—introduced briefly in the book's sixth chapter—names AI's most dangerous political capability: the capacity to simulate the common world without constructing it. Deepfakes, coordinated bot campaigns, AI-generated news articles designed to go viral—each can create the appearance that 'everyone is talking about this' when in fact the 'everyone' is partly or wholly manufactured. This synthetic joint attention is worse than joint attention's absence because it produces the behavioral effects of shared focus (mobilization, outrage, consensus) without the epistemic foundation (genuine encounter with a real common object) that makes those responses democratically legitimate. People act as if they have attended to the same reality. They have attended to different algorithmic constructions designed to produce the same response. The common world has been replaced by its simulation—and the simulation, because it is individually optimized, is more compelling than the reality it replaces.
The concept of the common world comes from Hannah Arendt's The Human Condition (1958), where it denotes the durable, shared reality of institutions, artifacts, and meanings that outlast individual lives and provide the stable ground for political action. Citton adapts the concept for the media age: the common world is not only institutions and artifacts but the shared attentional focus that keeps those institutions and artifacts meaningful. The common world requires that people attend, together, to the same things—and that the 'same things' are not algorithmically generated illusions but genuinely common objects. The adaptation is urgent because the media environment has shifted from scarcity (which forced shared attention) to abundance (which enables infinite personalization)—and abundance without governance destroys the commons.
Intersubjective construction. The common world is not the objective world but the shared world—constituted through joint attention, not given by physical reality.
Requires shared objects. Common world depends on materials that multiple minds can attend to simultaneously—newspapers, broadcasts, public events—which personalization eliminates.
Evaporation, not collapse. The common world does not end dramatically but thins gradually as the probability of shared reference declines, until strangers can no longer assume common ground.
Synthetic substitution. AI can simulate the common world—creating the appearance of shared focus through coordinated content—without constructing the genuine intersubjective ground that democracy requires.